2
\$\begingroup\$

I need to find a op-amp that will handle the high input voltages that my circuit requires. I'm not real familiar with them and not sure how to read their datasheets. I do have a LM741 but it looks like it can only handle +/- 15v. I need at least +/- 60v I think, but I have include the schematic that I used to test with. U1 is just LTSpice's ideal single-pole op-amp. Any suggestions on a particular op-amp or what I need to be looking for would be greatly appreciated.

http://img202.imageshack.us/img202/184/86py.png

Edit

I'm trying to use Vout to turn on the gate of a PMOS which is switching a variable voltage from a boost converter. V2 will come from boost converter and V1 will be what I subtracting from V2 so that the PMOS will cut on. The PMOS will be on when Source = Vout and off when Source = Source. The second schematic is the boost converter and the current configuration which uses a voltage divider to control the gate voltage, but I'm concerned that PMOS will not turn all the way on when the voltage is lowered all the way to 12v. I think that Vgs will just be to small using a voltage divider.

  • M2 is the PMOS
  • M3 is a NMOS used to control the voltage divider.
  • The two signals will be coming from Micro Controller.

http://imageshack.us/a/img22/4678/g16a.png

\$\endgroup\$
6
  • 3
    \$\begingroup\$ Is there a reason you can't use a resistor divider to bring the op-amp input voltages down to a lower range (This does tend to lose accuracy due to resistor tolerances)? What range of output voltages do you require? \$\endgroup\$ – The Photon Aug 9 '13 at 1:46
  • \$\begingroup\$ If not what The Photon said, I would seriously consider rewording your question to ask how to do whatever it is you're trying to do. An op amp with 60V rails will literally cost you hundreds of dollars, and they're not exactly a standard item. \$\endgroup\$ – Matt Young Aug 9 '13 at 1:53
  • \$\begingroup\$ @MattYoung, check this out: octopart.com/partsearch#search/requestData&q=ltc6090 \$\endgroup\$ – The Photon Aug 9 '13 at 1:58
  • \$\begingroup\$ @ThePhoton That is a cool little part. Definitely filed away to memory. I was thinking of the Apex stuff. apexanalog.com/wp-content/uploads/2012/10/PA89U_M.pdf \$\endgroup\$ – Matt Young Aug 9 '13 at 2:13
  • \$\begingroup\$ I'm just trying to subtract 10 volts from a voltage source ranging from 12v-60v. Voltage divider gets a fraction of the source so it would vary. Any ideas? \$\endgroup\$ – Wallace Aug 9 '13 at 2:42
2
\$\begingroup\$

I'm just trying to subtract 10 volts from a voltage source ranging from 12v-60v

Do you have any suggestion on how I could subtract 10 volts from source ranging from 12v-60v

I apologize for quoting comments from the OP rather than taking quotes directly from the question but it appears to me he is clearer in the his requirements in his comments.

Why don't you use a zener diode like this: -

enter image description here

If you then want to process the signal and you think that 50V is too high then the 4k7 can be made a potential divider such as a 1K and a 3k9. This will reduce the 50V to 10.2V and the 2V to 0.408V.

This won't be a perfectly accurate way of subtracting the voltage but you haven't said how accurate this has to be. You can get "zeners" that are very accurate in the form of shunt regulators - they only need a few tens of microamps to work.

\$\endgroup\$
1
  • \$\begingroup\$ Thank you this exactly the type of answer I was looking for. It doesn't have to be accurate at all because Vgs has a large range (+/- 30v) and will be completely turned on when Vgs = -10v. \$\endgroup\$ – Wallace Aug 11 '13 at 20:01
2
\$\begingroup\$

First, notice that in your circuit, the input voltage to your op-amp is only half of the V1 value. So if V1 is never more than 60 V, you really only need an op-amp capable of taking 30 V inputs. However, beware of "unusual" conditions like powering up and powering down --- you should be sure that the op-amp powers up before the input sources if you want to use this fact.

Second, op-amps with high voltage capability aren't that hard to find if you know where to look. I did a quick parametric search on linear.com and found several that will operate with more than 40 V between the rails. LTC6090 supports up to 140 V between the rails (+/- 70 V). It is, however, not a 25-cent jelly-bean part like the '741. I see pricing starting above $5 per part on Octopart.

Edit

Adapting my answer to a previous question, if I understand your overall problem, you maybe can do something like this.

schematic

simulate this circuit – Schematic created using CircuitLab

Here the comparator (OA1) never sees more than 6 V, so could be operated from 10 V supplies or so. You could do an even more extreme divider ratio to allow you to use a 5 V comparator, but the tolerances of the resistors build up more the lower the divider ratio. You'll probably want to use 1% or better resistors here even as I've drawn it.

The other thing to be careful about is looking for MOSFETs for M1 and M2 that will switch fast enough for your application. The value of R5 will also affect switching speed.

Edit 2 I show a resistor divider between M1 and M2 in response to your comment.

\$\endgroup\$
13
  • \$\begingroup\$ That is a lot of money to just subtract 10 volts from a variable Voltage source. Do you have any suggestion on how I could subtract 10 volts from source ranging from 12v-60v. \$\endgroup\$ – Wallace Aug 9 '13 at 2:40
  • \$\begingroup\$ I suspect you could build a simple but adequate difference amp from discrete transistors. But finding the right parts to do it and add up to less than $5 would take some time. \$\endgroup\$ – The Photon Aug 9 '13 at 2:44
  • \$\begingroup\$ See my recent answer on another question for a circuit that switches a PMOS at high voltage under control of a low voltage signal. \$\endgroup\$ – The Photon Aug 9 '13 at 2:46
  • \$\begingroup\$ Thank you for your help! Your post to another question help me out a ton. I have It working perfectly in LTSpice and should be cheaper than using the op-amp. \$\endgroup\$ – Wallace Aug 9 '13 at 3:19
  • \$\begingroup\$ Ok, I thought it was working but I'm still getting too high a Vgs. \$\endgroup\$ – Wallace Aug 9 '13 at 3:31
0
\$\begingroup\$

The LM741 Jellybean opamp has maximum power supply rails of (+/-)18v to 22 v depending on the particular version used (the LM741C has a maximum of 18v other flavors are 22v). It also states in the absolute maximum ratings that it has a maximum differential input voltage of 30v so this wont work in your application.

High voltage opamps are used. For example TI's OPA547T is rated for power supply rails up to +/- 30V and is rated for Common Mode voltage input ranges just shy (by 2 volts) of the power rails. When selecting an opamp for this application, you are going to want to look at the Common mode voltage range or differential input range on data sheets.

Another option is (again) TI's OPA445 which has a maximum differential signal input of 90 volts. It is important to note though, that to achieve such a great input range, the power supply voltage must be slightly larger than the desired input voltage. In the case of this chip, you must supply power at 10 volts higher than the maximum voltage you wish to measure (up to a differential of 100 volts). This is given again by the Common mode voltage range section of the data sheet.

To protect the inputs of the opamp you choose to use, you may want to clamp the inputs with Zener diodes (with the zener breakdown point selected slightly higher than the anticipated maximum input voltage) between the two inputs. This would prevent unanticipatedly high input voltages from causing damage to the opamp.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.